Visualize cafes on map with Folium

1. Coffee Bean shops on Korea map

Pinpoint Coffee Bean shops on the map

Cafes are very common in Korea, and many of them are franchise business. In this excercise I will visualize one of the franchise, Coffee Bean, with around 240 shops on the map.

Add district line on the map

To have a better idea of cafe distribution in different districts, I will add the district line.

2. Business Insight (ideas)

To answer the question: Is it a good choice to set up cafes in the selected location?

선택한 카페위치가 좋은 선택입니까?

1. Competition

Hypothesis: high competition may not be advantageous to Coffee Bean but moderate competitions indicate the coffee demand in the district. Areas with median number of cafes would be good choice of location.

I try to visualize other cafes in the neighborhood for a breif idea.

가설 : 높은 경쟁은 Coffee Bean에 유리하지 않을 수 있지만 적당한 경쟁은 해당 지역의 커피 수요를 나타냅니다. 카페수의 중간수준으로 좋은 것같습니다.

Brief 아이디어를 보여주기 위해 다른 카페 같은지역에서 시각화합니다.

For example, there is high cafe density in 서대문구 including coffee bean. 예를 들어 서대문구에는 커피 빈을 포함한 높은 카페 밀도가 있습니다.

Target Audience

Purpose to visit cafe: Cafe is usually the place for gatherings or a self-space for study/reading books...etc. The coffee price at Coffee Bean is considered as medium, targeting audience with some spending power.

Ideas for target segments:

--> Focus on office workers, University students, couples and friends as the core audience

--> Age between 20 - 45

--> Median income level

카페는 일반적으로 모임을하거나 책을 읽을 수있는 셀프 공간을 제공하는 곳입니다. 커피 가격은 중간 수준인데 약간 소비력을 가진 소비자을 대상으로합니다.

Target segment 아이디어 :

-> 직장인, 대학생, 커플 및 친구 = core

-> 20-45세

-> 중간 소득 수준

2. Office buidling, Shopping mall, Universities and Restaurants numbers

Hypothesis: office workers, University students and couples/friends tend to visit cafe more often.

--> Neighhoods with more offices, shopping malls, Universities (or restaurants) will be better choice.

가설 : 직장인, 대학생, 커플 / 친구가 카페를 더 자주 방문하는 경향이 있습니다.

-> 사무실, 쇼핑몰, 대학교 (또는 식당)이 더 많은 지역이 더 나은 선택이 될 것입니다.

3. Population density

Hypothesis: the more crowed area, teh higher demand of cafes

--> Explore the population density in each district

가설 : 사람 많은 지역 카페 수요 증가

-> 각 지역의 인구 밀도 탐색

4. Demographics within the area

Hypothesis: Age between 20 - 45 would be the target audience age levels

--> Explore the demographics of each district and check % population of aged 20-45, the higher the better

가설 : 20 세~45세 연령이 주요대상으로 합니다.

-> 인구 통계를 탐색하고 각지역의 20-45 세의 인구 비율을 확인합니다.

5. Income

Hypothesis: Targeting audience have some spending power for medium-priced coffee.

--> Find the area with median income level or above.

가설 : 고객은 야간 소비력이 있습니다.

-> 각지역 소득수준 분석하고 중간 소득 수준혹은 중간이상의 지역 찾으십시오.

3. K-means Clustering - location strategy

Identify the features which will have impact on cafe success. Group the districts into different clusters based on

  1. competitor numbers (no of other cafes)
  2. office/universities/shopping malls numbers
  3. population density
  4. core age group(20~45) %
  5. income level

카페위치 segment하십시오. 아래 data 구하고 여러 클러스터로 그룹화합니다.

  1. 경쟁자 수 (다른 카페 수)
  2. 사무실 / 대학교 / 쇼핑몰 수
  3. 인구 밀도
  4. 핵심 연령대 (20~45)%
  5. 소득 수준

To be continued... More data will be required for model preparation.

Conclusion

In this notebook, I visualized the franchise cafe Coffee Bean on the map using folium. By adding district line and other cafes on the same map, we get a better idea of the cafe competitions in the neighborwood. Obviously, there would be more cafes in the commercial area. I tried to hypothesize the target audience of franchise cafe and the project might be continued by validating the hypotheses with more data. Ultimately, we would be able to score the districts using K-means clustering model and aim to find out the best strategic locations for the franchise cafe.

Reference: